580 THE SENSES AND SENSOfi Y ORGANS. 



seen in optical section. This image is rose-coloured or green. 

 I have seen such an image in all the insects' eyes I have 

 examined in this way ; it is brilliant in the Diptera, which do 

 not exhibit the phenomenon of pseudo-pupils. 



The central pseudo-pupil is small, intensely black, and circular. 

 In this it differs entirely from the pseudo-pupils of the second 

 and third order, which are larger, dim, and irregular in form. 

 All the pseudo-pupils appear to be beneath the cornea. 



Pseudo-pupils of the Second Order. These are six in number ; 

 they are twice as large as the central pseudo-pupil, and gray in 

 colour. I have little doubt these are diffraction images ; they 

 correspond with the angles of a hexagon circumscribed around 

 the central pseudo-pupil as its centre. 



Pseudo-pupils of the Third Order. These are approximately 

 twice the diameter of those of the second order, more irregular, 

 and grayer. They are t\velve in number, but generally only 

 two or three are visible at once, as they fall outside the com- 

 pound cornea. The whole circle can never be seen at the 

 same time, but they may be brought into view in succession by 

 causing the central pseudo-pupil to fall on the edge of the 

 cornea, and to travel round it by moving the insect. 



The pseudo-pupils of the second and third order correspond 

 with the diffraction images produced by a hexagonal grating. 



Exner gives a totally different explanation of the pseudo- 

 pupils of the second and third order, and ascribes them to the 

 manner in which the pigment of the chamber is arranged. 

 For his explanation I must refer the reader to his mono- 

 graph [252]. 



It will be seen that whatever the true explanation of the 

 pseudo-pupils may be, the phenomenon throws no light on the 

 mariner in which images are formed on the retina, although the 

 bright image of the retinula seen with the ophthalmoscope 

 and microscope combined tends towards the verification of the 

 hypothesis I have adopted in these pages, since it shows that a 

 real image of the retinula can be produced in front of the eye, 

 just as a real image of the vertebrate retina can be seen with 

 the ophthalmoscope. 



